TCR Revision Generates Functional CD4+ T Cells J. Scott Hale, Maramawit Wubeshet and Pamela J. Fink This information is current as J Immunol 2010; 185:6528-6534; Prepublished online 22 of September 23, 2021. October 2010; doi: 10.4049/jimmunol.1002696 http://www.jimmunol.org/content/185/11/6528 Downloaded from

References This article cites 36 articles, 21 of which you can access for free at: http://www.jimmunol.org/content/185/11/6528.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 23, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

TCR Revision Generates Functional CD4+ T Cells

J. Scott Hale, Maramawit Wubeshet, and Pamela J. Fink

CD4+Vb5+ peripheral T cells in C57BL/6 mice respond to encounter with a peripherally expressed endogenous superantigen by undergoing either deletion or TCR revision. In this latter process, cells lose surface Vb5 expression and undergo RAG-dependent rearrangement of endogenous TCRb genes, driving surface expression of novel TCRs. Although postrevision CD4+Vb52TCRb+ T cells accumulate with age in Vb5 transgenic mice and bear a diverse TCR Vb repertoire, it is unknown whether they respond to homeostatic and antigenic stimuli and thus may benefit the host. We demonstrate in this study that postrevision cells are functional. These cells have a high rate of steady-state homeostatic proliferation in situ, and they undergo extensive MHC class II-dependent lymphopenia-induced proliferation. Importantly, postrevision cells do not proliferate in response to the tolerizing superantigen, implicating TCR revision as a mechanism of tolerance induction and demonstrating that TCR-dependent activation of postrevision cells is not driven by the transgene-encoded receptor. Postrevision cells proliferate extensively to commensal b bacterial Ags and can generate I-A –restricted responses to Ag by producing IFN-g following Listeria monocytogenes challenge. Downloaded from These data show that rescued postrevision T cells are responsive to homeostatic signals and recognize self- and foreign peptides in the context of self-MHC and are thus useful to the host. The Journal of Immunology, 2010, 185: 6528–6534.

eveloping T cells within the thymus undergo rigorous tional selective force that maintains useful components of the peri- selection to ensure the usefulness and safety of mature pheral T cell repertoire (reviewed in Ref. 2).

T cells bearing rearranged ab TCRs. Thymocytes at the Mechanisms in the lymphoid periphery exist to keep in check http://www.jimmunol.org/ D2 2 CD4 CD8 double-negative stage undergo RAG-mediated TCRb self-reactive peripheral T cells that elude negative selection in gene rearrangement, and generation of a functional TCRb-chain the thymus. In Vb5 transgenic (Tg) mice, Vb5+CD4+ T cells in induces proliferation and CD4 and CD8 coreceptor expression. At the lymphoid periphery encounter an endogenous superantigen the CD4+CD8+ double-positive stage, a second wave of RAG- encoded by a defective mouse mammary tumor virus (Mtv) and dependent rearrangement, this time at the TCRa locus, gener- are tolerized to this self-Ag by several means. Mtv superantigen ates rearrangements that encode TCRa-chains that pair with the drives deletion of most CD4+ T cells, resulting in a severe age- already available TCRb-chains, and these TCRab pairs are tested dependent decline in the CD4/CD8 ratio (3, 4). Mtv drives some for interactions with self-MHC molecules. Useful affinity for self- cells to undergo TCR revision, through which CD4+ T cells lose MHC class II (MHC II) and class I molecules drives positive surface Vb5 expression, re-express RAG, and undergo TCRb by guest on September 23, 2021 selection, upregulation of TCR surface expression, and commit- gene rearrangement, resulting in the age-dependent accumulation ment to the CD4 and CD8 T cell lineages, respectively. Insuf- of a rescued population of postrevision CD4+Vb52TCRb+ T cells ficient interactions with MHC result in death by neglect. Positively (5, 6). The expression of RAG (5, 7–18) and the presence of TCR selected CD4 single-positive and CD8 single-positive thymocytes Vb-(DJ)b or Va-Da recombination intermediates (5, 7, 11, 13–17) test their TCRs against a broad array of self-MHC–self-peptide in peripheral T cells, as well as a blockade in the generation of ligands for self-reactivity, and those cells that recognize self are postrevision CD4+Vb52TCRb+ T cells when Rag2 is condition- removed from the repertoire by apoptosis (reviewed in Ref. 1). ally deleted in postthymic mature Vb5+ T cells (19), unequivo- After thymic exit, basal TCR signaling in the lymphoid periphery cally identify TCR revision as a process that targets mature CD4+ promotes survival and homeostatic proliferation and is an addi- Vb5+ cells in the lymphoid periphery for RAG-dependent gen- eration of new TCRs (reviewed in Ref. 20). Postrevision T cells express a diverse TCRb repertoire with TCRs characterized by shortened CDR3 loops resulting from fewer N nucleotides (18). Department of Immunology, University of Washington, Seattle, WA 98195 Given the specialized selecting environment of the thymus, it is Received for publication August 9, 2010. Accepted for publication September 24, unknown whether cells generated through TCR revision in the 2010. lymphoid periphery have been selected for appropriate T cell 2 This work was supported by the National Institutes of Health (Grant RO1 AG 13078 function that could benefit the host. Although postrevision Vb5 to P.J.F.), the Cancer Research Institute’s Predoctoral Emphasis Pathway in Tumor T cells proliferate in response to TCR cross-linking (5) and con- Immunology Program (to J.S.H.), and the National Cancer Institute Basic and Cancer Immunology (Grant T32CA09537 to J.S.H.). tain a population of IL-17–producing cells following PMA plus The content of this work is solely the responsibility of the authors and does not ionomycin stimulation (21), their homeostatic potential and im- necessarily represent the official views of the National Institutes of Health, the munocompetence remain untested. In this study, we assessed the Cancer Research Institute, or the National Cancer Institute. ability of postrevision cells to respond to homeostatic factors in- Address correspondence and reprint requests to Dr. Pamela J. Fink, University of cluding self-MHC and foreign Ag following bacterial challenge of Washington, 1959 NE Pacific Street, Health Sciences Building I607H, Seattle, WA 98195. E-mail address: pfi[email protected] Vb5 Tg mice. Our data indicate that postrevision T cells require Abbreviations used in this paper: B6, C57BL/6; iono, ionomycin; LIP, lymphopenia- TCR–MHC interactions for maximal lymphopenia-induced pro- induced proliferation; LLO, listeriolysin O; MHC II, MHC class II; MLN, mesenteric liferation (LIP) and the generation of Ag-specific self-MHC– lymph node; Mtv, mammary tumor virus; nonTg, nontransgenic; Tg, transgenic; restricted effector T cell responses to bacterial Ag. Additionally, unstim, untreated. postrevision cells possess no residual reactivity to Mtv super- Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 antigen, showing that TCR signaling occurs through the revised www.jimmunol.org/cgi/doi/10.4049/jimmunol.1002696 The Journal of Immunology 6529

TCR and not the transgene-encoded receptor, further defining TCR cultured for 3 d at 37˚C and 7% CO2, followed by surface staining and flow revision as a mechanism of peripheral T cell tolerance. cytometric analysis. Listeria monocytogenes infections Materials and Methods Wild-type L. monocytogenes was grown in brain–heart infusion broth and Mice measured by OD (A600) at midlog growth phase. The culture was diluted in sterile PBS and 200 ml containing 2000 CFU injected into the lateral tail Vb5 Tg and nontransgenic (nonTg) littermates on the C57BL/6 (B6) vein. background were bred under specific-pathogen free conditions at the University of Washington (Seattle, WA). Mice on the B6 background carry Intracellular cytokine staining the endogenous proviral genes encoding Mtv-8, -9, -17, and -30. B6 Ly5.2+ a b + 6 and B6.SJL (B6.SJL-Ptprc Pepc /BoyJ) Ly5.1 mice and their F1 gener- To detect IFN-g–producing T cells from uninfected mice, 2 to 3 3 10 ation (Ly5.1+Ly5.2+) were purchased from The Jackson Laboratory (Bar splenocytes were incubated in 96-well plates for 6 h at 37˚C in RP10 in the Harbor, ME) or bred in-house. B6.PL-Thy1a/CyJ (Thy-1.1+) mice were presence of monensin (GolgiStop, BD Pharmingen) either in the absence 2 2 kindly provided by M.K. Kaja (University of Washington). CD4 / (22) or presence of 0.7 mM ionomycin plus 50 ng/ml PMA (both from Sigma- 2 2 and I-Ab / (23) mice on the B6 background were originally purchased Aldrich). For detection of L. monocytogenes-specific CD4+ T cells, 2 2 2 2 from The Jackson Laboratory. Rag2 / (hereafter Rag / ) mice on the B6 splenocytes from mice infected 7 d previously were incubated in 96-well background were provided by D. Stetson (University of Washington). To plates for 5 h at 37˚C in RP10 in the presence of monensin and either in the induce lymphopenia in recipient mice, whole-body sublethal irradiation of absence or presence of 1 mg/ml I-Ab–restricted peptide listeriolysin O 650 rad was administered 1 d before cell transfer. For antibiotic treatment (LLO) (NEKYAQAPNVS) from Invitrogen (Carlsbad, CA). Postin- 2 2 190–201 of Rag / mice, drinking water containing 1 g/ml neomycin trisulfate, 1 g/ cubation, cells were surface stained, followed by intracellular staining with ml ampicillin (both from Sigma-Aldrich, St. Louis, MO), 1 g/ml polymixin allophycocyanin-conjugated anti–IFN-g using the Cytofix/Cytoperm kit and B sulfate (Life Technologies, Rockville, MD), and 50 mg/ml ciprofloxacin protocol (BD Biosciences). Downloaded from (Claris Lifesciences, Ahmedabad, India) was administered beginning 6 d before cell transfer and made fresh every 6 d. All experiments were conducted in accordance with the University of Washington Institutional Results Animal Care and Use Committee. Elevated steady-state proliferation by postrevision T cells Cell preparation and flow cytometry To determine whether postrevision CD4+Vb52TCRb+ T cells are

Single-cell suspensions were prepared from spleen and mesenteric lymph functional, we first examined their steady-state proliferation. Aged http://www.jimmunol.org/ nodes (MLNs), and RBC were removed from spleen by water lysis. For flow Vb5 Tg and nonTg mice were administered BrdU for 5 d and CD4+ cytometry, cells were surface stained in HBSS containing 1% BSA. FcRs T cells analyzed for BrdU incorporation. Postrevision CD4+Vb52 were blocked using anti-CD16/32 (2.4G2; BD Pharmingen, San Diego, CA) T cells in Vb5 Tg mice express high levels of surface CD44 (Fig. followed by staining with FITC, PE, PerCP-Cy5.5, PE-Cy7, allophyco- cyanin, and allophycocyanin-Alexa Fluor 750 or -Alexa Fluor 780 1A). Therefore, we compared the basal proliferation of postrevision + 2 + + high + fluorochrome-conjugated Abs and, in some cases, biotinylated Abs fol- CD4 Vb5 cells to that of CD4 Vb5 CD44 CD4 T cells from lowed by allophycocyanin-conjugated streptavidin. Abs were purchased Vb5Tg mice and to that of Vb52CD44high and Vb5+CD44high from BD Biosciences (San Jose, CA) or eBioscience (San Diego, CA) and CD4+ T cells from nonTg mice (Fig. 1). Postrevision Vb52 cells included mAbs recognizing mouse CD4 (RM4-5), CD44 (IM7), CD45.1/ Ly5.1 (A20), CD45.2/Ly5.2 (104), CD62L (MEL-14), Thy-1.2 (53-2.1), underwent a significantly higher rate of steady-state proliferation, by guest on September 23, 2021 pan-TCRb (H57-597.13), and Vb5 (MR9-4). Flow cytometry data were with 40–50% incorporating BrdU during the 5-d labeling period, 2 collected using a FACSCanto (BD Biosciences) and analyzed with FlowJo compared with 25–30% of Vb5 CD44high CD4+ T cells in nonTg software (Tree Star, Ashland, OR). Cell sorting was done using a FAC- mice and Vb5+CD44high cells in Vb5 Tg mice (Fig. 1B,1C). This SAria (BD Biosciences). more extensive proliferation of postrevision T cells was observed in Cell enrichments, CFSE labeling, and cell transfers both spleen and MLNs (Fig. 1C). Untouched CD4 T cell enrichments were performed using an EasySep Postrevision cells undergo extensive MHC II-dependent LIP + Negative Selection Mouse CD4 T cell Enrichment Kit (StemCell Tech- + nologies, Vancouver, British Columbia, Canada). For cell transfers, en- Homeostatic proliferation of CD4 T cells is regulated by several riched or sorted cells were CFSE labeled at a concentration of 50 3 106 factors, including TCR–MHC/self-peptide interactions and cyto- cells/ml in 5 mM CFSE for 10 min at 37˚C. CFSE labeling was arrested by kines. Transfer of T cells into recipient mice made acutely lym- adding cold HBSS containing 1% BSA. Labeled cells were washed and phopenic through irradiation provides a sensitive model to assess injected into the lateral tail vein in a 200 ml volume of HBSS. the requirements for homeostasis of T cell subsets (24). In acutely In situ BrdU labeling lymphopenic recipients, transferred CD4+ T cells undergo LIP, Mice were injected i.p. with 0.8 mg BrdU (Sigma-Aldrich) in a 200 ml typically revealing a subpopulation that undergoes rapid cell pro- volume, followed by administration of sterile drinking water containing liferation driven by contact with self-MHC–peptide complexes 0.8 mg/ml BrdU, made fresh and changed daily for 5 d. For flow cyto- and a subpopulation that undergoes slower cell division that is metric detection of BrdU, cells were surface stained, followed by intra- IL-7 dependent (24). To further characterize the functionality of nuclear staining with FITC-conjugated anti-BrdU (BD Biosciences) using the BrdU Flow Kit and protocol (BD Biosciences). postrevision T cells in vivo, we first tested their ability to respond to homeostatic signals by assessing their capacity to undergo LIP In vitro T cell activation following transfer of CFSE-labeled CD4+ T cells from Vb5Tg A mix containing an equal number of congenically marked nonTg and Vb5 mice into sublethally irradiated syngeneic recipient mice. Donor Tg splenocytes were CFSE labeled. A total of 2 3 106 CFSE-labeled CD4+ T cells were easily identifiable within recipient spleens by mixed splenocytes were plated in a 24-well plate either in the absence congenic marker staining and were further gated into CD4+Vb52 3 5 or presence of 1–3 10 irradiated (4300 rad) PN53HI B lymphoma cells and CD4+Vb5+ donor populations (Fig. 2A). Postrevision CD4+ that express high levels of Mtv superantigen and drive measurable pro- 2 ∼ liferation of Vb5+ T cells in vitro. PN53HI cells were isolated from tumor- Vb5 cells underwent substantial LIP, with 60% of cells di- prone P53+/2 perforin null mice and were kindly provided by Mark J. viding during this 8-d time period. In addition to a more slowly Smyth (Peter MacCallum Cancer Centre, East Melbourne, Victoria, Aus- dividing subset of cells, .30% of CD4+Vb52 cells were rapid tralia). Alternatively, CFSE-labeled splenocytes were stimulated with 5 ng/ dividers, having undergone more than five cell divisions (Fig. 2A). ml anti-CD3ε (clone 145-11C) plus 1 mg/ml anti-CD28 (clone 37.51), both + from BD Pharmingen, in a total volume of 2 ml RP10 (RPMI 1640 with Thus, postrevision CD4 T cells respond to homeostatic factors 10% FBS, 10 mM HEPES, 4 mM L-glutamine, 100 U/ml penicillin, 100 and proliferate in a lymphopenic environment to a greater extent mg/ml streptomycin, 50 mM 2-ME, and 50 mg/ml gentamycin). Cells were than do transferred CD4+Vb5+ cells. 6530 POSTREVISION T CELLS ARE FUNCTIONAL Downloaded from

FIGURE 2. MHC II and other factors drive extensive LIP of post- revision CD4+Vb52 T cells. Total of 3 3 106 CFSE-labeled CD4+ T cells enriched from pooled spleen and lymph nodes of aged Vb5 Tg mice were transferred into the indicated congenic recipients that had been sublethally http://www.jimmunol.org/ irradiated (650 rad) 1 d previously to induce lymphopenia. Donor CD4+ T cells were analyzed 8 d later for the extent of proliferation as measured by CFSE dilution. For CFSE histograms, numbers represent the percent of gated cells that are undivided (upper right panel), have undergone at least one division (upper left panel), or have divided more than five times (lower left panel). A, Analysis of B6 recipient spleen with gate indicating Ly5.1+ donor CD4+ T cells (top left panel) and Vb5 expression of gated donor CD4+Ly5.1+ T cells (top right panel). CFSE analysis of donor CD4+Vb52 + +

(bottom left panel) and donor CD4 Vb5 (bottom right panel) T cells. by guest on September 23, 2021 Data are representative of two independent experiments analyzing a total of four recipients. B, Analysis of donor CD4+Vb52 (left panels) and CD4+ Vb5+ (right panels) cells posttransfer into CD42/2 (MHC II+ lymphopenic FIGURE 1. Postrevision T cells show an elevated steady-state pro- control) and I-Ab2/2 recipients. Data are representative of three in- liferation in situ. Vb5 Tg and nonTg mice 25–30 wk of age were given dependent experiments and a total of five to six recipients of each geno- BrdU by i.p. injection, followed by a 5-d administration of BrdU in the type. The percent of donor Vb52 postrevision cells that had fully diluted drinking water. Spleen and MLN cells were isolated and stained for surface CFSE was 36.1 6 3.8 in CD42/2 recipients and 7.1 6 1.6 in I-Ab2/2 markers and BrdU incorporation. A,Vb5 and CD44 analysis of CD4+ 2 recipients. p = 0.0001 as determined using a two-tailed nonparametric gated splenocytes, showing further gating of CD44high Vb5 and Vb5+ Student t test. T cells from nonTg and Vb5 Tg mice. B, Representative BrdU in- corporation in CD4+CD44high Vb52 and Vb5+ splenocytes from nonTg and Vb5 Tg mice. C, Charts indicate the percent BrdU+ for CD4+CD44high Postrevision cells undergo LIP in competition with nonTg gated populations from spleen and MLNs of all nonTg (filled circles) and memory phenotype cells Vb5 Tg (open circles) mice analyzed. Data are compiled from three in- To determine whether postrevision cells can undergo LIP even in dependent experiments (Vb5 Tg mice: n = 4; nonTg mice: n = 5). Bars + + + high competition with nonTg CD4 T cells with a similar CD44 phe- represent the mean percent of BrdU cells of the CD4 CD44 pop- high + ulations. p values were calculated using a two-tailed nonparametric Stu- notype, an equal number of congenically marked CD44 CD4 dent t test. Significant p values are shown; n.s. indicates p values .0.05. T cells from aged nonTg and Vb5 Tg mice were cotransferred into acutely lymphopenic hosts. CD4+ T cells from donor nonTg and Vb5 Tg mice that were sorted to obtain CD62L2CD4+ T cells had To determine whether MHC-dependent TCR signaling con- indistinguishably high surface CD44 expression (Fig. 3A). An equal tributes to the extensive LIP of postrevision T cells, CFSE-labeled number of sorted CD4+CD44high nonTg and Vb5 Tg cells were + Vb5 Tg CD4 T cells were transferred to sublethally irradiated mixed, CFSE labeled, and transferred into congenically marked 2/2 2/2 recipient I-Ab or similarly lymphopenic control CD4 B6 irradiated recipient mice. A similar proportion of postrevision cells mice. Substantial proliferation of postrevision cells occurred in from Vb5 Tg donors underwent cell division compared with CD4+ both types of recipients in 8 d; however, the rapidly dividing Vb52CD44high cells from nonTg mice and were not outcompeted 2/2 population of postrevision cells was absent in I-Ab recipients by them for proliferation (Fig. 3B) or recovery (data not shown), 2/2 compared with control CD4 hosts (Fig. 2B). As expected, the although a reduced population of rapidly dividing cells was ob- + + smaller, rapidly proliferating population of CD4 Vb5 cells was served among the postrevision population (Fig. 3B). also dependent on the presence of MHC II molecules (Fig. 2B). These experiments define MHC II as a critical factor for maximal Postrevision T cells are tolerant to Mtv superantigen homeostatic proliferation of postrevision CD4+ T cells under Previous attempts to use cytokine secretion or proliferation to lymphopenic conditions. measure Vb5+ T cell reactivity to Mtv superantigens on normal The Journal of Immunology 6531

FIGURE 4. Postrevision T cells are tolerant to Mtv superantigen. Downloaded from Splenocytes from aged Ly5.1+ nonTg and Ly5.2+ Vb5 Tg mice were mixed, CFSE labeled, and cultured for 3 d in the presence of either 3 3 105 irradiated Mtv-overexpressing PN53HI tumor cells (A) or soluble anti-CD3 plus anti-CD28 (B). Three days later, cells were analyzed for proliferation FIGURE 3. Postrevision T cells undergo LIP when competing against as measured by CFSE dilution. Numbers in CFSE histograms represent the TCR nonTg effector/memory phenotype cells. CD4+ T cells from 25–30- percent of gated cells that have undergone at least one division. Data are + + + wk-old Thy-1.2 nonTg (Ly5.2 ) and Vb5 Tg (Ly5.1 ) mice were enriched representative of four independent experiments. http://www.jimmunol.org/ by negative selection, stained with CD62L, and sorted to isolate untouched CD62L2CD4+CD44high T cells. An equal number of sorted CD4+ Tg and nonTg cells were mixed, labeled with CFSE, and a total of 2 3 106 CD4+ Postrevision T cells respond to commensal bacteria in 2 T cells adoptively transferred into sublethally irradiated Thy-1.1+Thy-1.2 chronically lymphopenic mice B6 recipient mice. Thy-1.2+ donor CD4+CD44high splenocyte populations were analyzed for the extent of proliferation as measured by CFSE dilution. To determine whether postrevision T cells can proliferate in re- A, CD62L and CD44 analysis of pre- and postsort CD4+ T cells from nonTg sponse to microbial Ags, we adoptively transferred CFSE-labeled + and Vb5 Tg mice. B, CFSE analysis of donor CD44high nonTg and Vb5Tg CD4 T cells from aged Vb5 Tg mice into chronically lympho- 2 2/2 CD4+Vb5 (left panels) and CD4+Vb5+ (right panels) cells 8 d posttransfer penic Rag recipient mice. These mice support extensive pro- into sublethally irradiated recipients. Numbers in CFSE histograms repre- liferation of transferred T cells driven by an exaggerated burden of by guest on September 23, 2021 sent the percent of gated cells that are undivided (upper right panel), have microbial Ags that results from their immunodeficient state (27). undergone at least one division (upper left panel), or have divided more than To assess the relative contribution of bacterial Ags to the pro- five times (lower left panel). Data are representative of two independent 2/2 2 liferation of the transferred T cells, a cohort of Rag recipients experiments for a total of four recipients. The percent of donor Vb5 cells treated with a mixture of antibiotics was included. Six days fol- that had undergone at least one cell division was 62.3 6 4.0 for nonTg cells + 2/2 6 2 lowing transfer, donor CD4 T cells comprised 2 to 3% of Rag and 72.1 4.2 for Vb5 postrevision cells. p = 0.1391 as determined using + a two-tailed nonparametric Student t test. recipient spleen cells and 50% of cells in the MLN; the donor CD4 T cell frequency was dramatically reduced in antibiotic-treated recipient mice (Fig. 5A). Postrevision Vb52 cells accumulated APCs have been unsuccessful, possibly due to the poor expression preferentially compared with Vb5+ cells in Rag2/2 recipients, of the relevant Mtvs and the weak nature of this ligand for TCR whereas antibiotic treatment resulted in a Vb52 to Vb5+ cell ratio Vb5 (25, 26). To measure the Mtv reactivity of postrevision similar to that of the inoculum (Fig. 5B and legend). In untreated T cells, we made use of PN53HI B lymphoma cells that express recipients, donor Vb52 and Vb5+ cells divided extensively, with + + very high levels of Mtv, driving proliferation of Vb5 CD4 and .80% of cells having completely diluted their CFSE, whereas + CD8 T cells in vitro and in vivo (Fig. 4A and data not shown). To proliferation was greatly retarded in antibiotic-treated recipients determine whether postrevision T cells are Mtv tolerant, CFSE- (Fig. 5C). Donor Vb52 postrevision T cells were recovered at high labeled splenocytes mixed from nonTg and Vb5 Tg mice were frequencies in the MLN and to a lesser degree in the spleen of cultured in the presence of irradiated PN53HI tumor cells and the untreated recipients. The recovery of postrevision cells compared extent of their proliferation measured 3 d later. Despite driving with that of Vb5+ T cells was ∼5–7-fold higher in the spleen and + + proliferation of CD4 Vb5 nonTg and Vb5 Tg T cells, PN53HI 11–15-fold higher in the MLN of Rag2/2 recipients (Fig. 5D). tumor cells induced no proliferative response in either postrevision Antibiotic treatment of recipient mice resulted in a 700-fold re- + 2 + 2 CD4 Vb5 cells or CD4 Vb5 cells from nonTg mice (Fig. 4A). duction in donor CD4+Vb52 T cell recovery in the MLN and a 40- Thus, postrevision cells no longer recognize the Mtv tolerogen fold reduction in the spleen (Fig. 5D). Together, these data dem- that directed them into the TCR revision pathway. These cells are onstrate that postrevision cells, and to a lesser degree CD4+Vb5+ functional, however, as judged by their rapid proliferation in re- cells, proliferate and accumulate extensively in response to mi- sponse to anti-CD3 plus anti-CD28 stimulation (Fig. 4B). These crobial Ags in Rag2/2 recipient mice. data designate TCR revision as a tolerance pathway, generating + functional postrevision cells. In addition, TCR-dependent acti- Effector function of postrevision CD4 T cells vation/function of postrevision T cells likely comes through the To assess the immunocompetence of postrevision T cells, we first newly generated receptor and not through any residual expression determined whether they produce IFN-g. CD4+ T cells from and signaling via the transgene-encoded Vb5 TCR. aged nonTg and Vb5 Tg mice were stimulated with PMA plus 6532 POSTREVISION T CELLS ARE FUNCTIONAL Downloaded from

FIGURE 6. Postrevision T cells include IFN-g–producing cells. Splenocytes from 25–34-wk-old nonTg and Vb5 Tg mice were stimulated with PMA plus ionomycin (iono) for 6 h or left untreated (unstim), fol- lowed by surface CD4 and Vb5 and intracellular IFN-g staining. A, Representative Vb5 and IFN-g analysis of CD4+ T cells. B, Chart indicates the percent of stimulated Vb52 and Vb5+ CD4+ T cells from nonTg and http://www.jimmunol.org/ Vb5 Tg mice that produce IFN-g. Bars represent the mean percent of each indicated cell population with error bars indicating the SD (Vb5 Tg: n =3; nonTg: n = 3). No significant differences were obtained using a two-tailed nonparametric Student t test.

postinfection, Ag-specific CD4+IFN-g+ T cells were identified by culturing splenocytes from infected mice with the immunodo- b minant I-A –restricted Listeria epitope (LLO190–201 peptide), fol- by guest on September 23, 2021 lowed by intracellular cytokine staining (Fig. 7A). Postrevision 2 + Vb5 cells comprised .95% of the LLO190–201-specific CD4 FIGURE 5. Postrevision T cells undergo rapid proliferation to bacterial IFN-g+ population in Vb5 Tg mice (Fig. 7B). Similar to Ag- Ag following transfer into chronically lymphopenic recipients. Total of specific CD4+ T cells from infected nonTg mice, IFN-g+ LLO - 2.6 3 106 CFSE-labeled Ly5.1+ CD4+ T cells (composed of 12% Vb52 190–201 specific postrevision cells exhibited decreased levels of surface postrevision cells and 88% Vb5+ cells) enriched from pooled spleen and lymph nodes of aged Vb5 Tg mice were transferred into chronically TCRb, resulting from contact with specific peptide during the 5-h lymphopenic Rag2/2 recipient mice that were either untreated or trea- incubation (Fig. 7C). Thus, postrevision cells can recognize Ag in ted (+ Abx) with a mixture of antibiotics in their drinking water beginning a self-MHC–restricted manner and become functional cytokine- 6 d prior to adoptive transfer. Donor CD4+ T cells were analyzed 6 producing effector cells in response to a bacterial challenge. d posttransfer for the extent of proliferation as measured by CFSE dilution and the percent recovery of each population. A, Analysis of live gated Discussion splenocytes and MLN cells with gate indicating Ly5.1+ donor CD4+ 2 In this study, we show that postrevision CD4+Vb5 T cells gen- T cells. B, CD4 and Vb5 analysis of donor Ly5.1+CD4+ gated cells from the MLN. C, CFSE analysis of donor CD4+Vb52 (left panels) and CD4+ erated through peripheral TCR revision are capable of undergoing Vb5+ (right panels) T cells from the MLN. Numbers within CFSE his- homeostatic proliferation. Previous studies defining the homeo- + high 2 tograms represent the percent of gated cells that have fully diluted CFSE static requirements of memory phenotype CD4 CD44 CD25 (left panels) or have undergone 0–4 cell divisions (right panels). D, Charts T cells (which may or may not contain memory cells formed indicate the percent recovery of donor Vb52 and Vb5+ CD4+ T cells in through encounter with specific foreign Ag) upon transfer into spleen and MLN. acutely lymphopenic recipients have revealed a heterogeneous population of rapidly dividing cells that require MHC II for their extensive proliferation and a population that does not require ionomycin for 5 h and found to include similar frequencies of MHC II and relies on IL-7 to drive slow cell division (28). In con- IFN-g–producing cells (Fig. 6A). Comparison of Vb52 and Vb5+ trast, Ag-specific CD4+CD44high memory cells generated through CD4+ T cells from nonTg and Vb5 Tg mice indicated that encounter with a known viral Ag only underwent the slower, IL-7– postrevision cells and Vb5+ and Vb52 CD4+ T cells from age- dependent LIP (28). In this study, we show that postrevision CD4+ 2 matched nonTg mice have similar frequencies of IFN-g+ cells Vb5 T cells undergo more extensive LIP in irradiated recipients (Fig. 6A,6B). Thus, the postrevision T cell pool contains func- than their Vb5+ counterparts. This difference in proliferation may tional cytokine-producing cells that may contribute to host im- have resulted from the Mtv-dependent anergy that characterizes munity. the CD4+Vb5+ T cell population in B6 mice (4). Importantly, the To test whether postrevision cells are self-MHC restricted, aged most extensive LIP of postrevision cells requires the presence of Vb5 Tg mice were infected with L. monocytogenes. Seven days MHC II molecules, because the rapidly dividing population was The Journal of Immunology 6533

FIGURE 7. Postrevision Vb52 T cells pro- duce IFN-g in a primary response to Listeria infection. Aged nonTg and Vb5 Tg mice were infected with L. monocytogenes, and 7 d later, CD8-depleted splenocytes were cultured for 5 h with or without LLO190–201 peptide, followed by surface staining for CD4, Vb5, and TCRb and intracellular staining for IFN-g. A, Intracellular IFN-g staining of CD4+ gated splenocytes. B, Vb5 expression by total CD4+ and Ag-specific CD4+IFN-g+ gated splenocytes. C, Surface pan- TCRb expression by CD4+ gated and Ag-specific CD4+IFN-g+ gated splenocytes. absent following transfer to irradiated I-Ab2/2 mice. In addition Previously, our laboratory has shown that the Foxp3+ regulatory to the rapidly proliferating population, we also noted a slower T cell lineage is excluded from the postrevision population and dividing population of postrevision T cells that is not dependent that TCR revision likely skews the postrevision T cell compart- on the presence of MHC and may instead be driven by IL-7. Our ment in favor of IL-17–producing Th17 cells (21). In this study, comparison of cotransferred postrevision and effector/memory we observed that the postrevision compartment in aged Vb5Tg 2 nonTg CD44highCD62L cells demonstrates that postrevision cells mice includes a population that readily produces IFN-g when Downloaded from compete effectively with effector/memory CD4+ T cells from nonTg stimulated directly ex vivo with PMA plus ionomycin; however, mice for homeostatic factors. In summary, similar to CD4+CD44high the frequency of IFN-g+ cells is similar to that found in the Vb5+ memory phenotype cells from nonTg B6 mice, postrevision cells population and in the CD4+ T cell population in age-matched transferred into acutely lymphopenic irradiated hosts require the nonTg mice. Thus, TCR revision does not preferentially skew the presence of MHC II for the most rapid proliferation and other fac- postrevision T cell compartment toward IFN-g production.

tors (likely IL-7) for slower proliferation. Postrevision T cells proliferate extensively to bacterial Ags http://www.jimmunol.org/ Unlike memory phenotype cells from nonTg mice, the in situ following transfer to Rag2/2 mice and produce IFN-g in response steady-state proliferation by postrevision T cells (as measured by to L. monocytogenes challenge of Vb5 Tg mice. Thus, post- BrdU labeling) appears to be more rapid than that of the other revision T cells are selected, possessing the ability to recognize CD44high cell populations analyzed (Fig. 1). Several possibilities foreign Ag in the context of self-MHC. It is unknown whether the could explain such rapid basal proliferation. First, postrevision revised TCRb-chains pair with the TCRa-chains originally gen- cells exist in a CD4-lymphopenic setting, and thus, their high erated and selected within the thymus of the Vb5 Tg mouse or basal proliferation may be driven by the increased abundance of whether TCR revision at the TCRb locus is accompanied by homeostatic cytokines and self-MHC ligand compared with that TCRa locus rearrangement to form a functional, revised TCR that available to CD44high cells in lymphoreplete nonTg B6 mice. can generate these Ag-specific responses. In either case, the short- by guest on September 23, 2021 However, the steady-state proliferation of postrevision cells is also ened CDR3 loops of revised TCRb-chains may be critical for increased compared with that of Tg CD44highVb5+ cells sharing endowing the revised TCRab pair with the ability to interact with the same environment, suggesting that lymphopenia is not the MHC molecules in a useful way, allowing for more efficient se- only factor that determines their increased proliferation. A second lection, an interpretation suggested by the increased efficiency of possibility is that extrathymic selection of postrevision T cells positive selection of TdT2/2 thymocytes (35). Our findings sug- generates a more self-reactive TCR repertoire. The decreased gest that MHC II+ cells in the lymphoid periphery, perhaps under number of N nucleotides and resultant shortened TCRb CDR3 re- specific conditions, possess the ability to select the TCRs gener- gions of revised TCRs (18) may endow postrevision TCRs with ated through TCR revision, allowing postrevision T cells to see a more promiscuous peptide recognition and an increased affinity both MHC–self-peptide and MHC–foreign peptide ligands. The for MHC II a helices, similar to that characterizing TCRs of germinal center location of RAG+ revising T cells may provide a neonatal T cells (29–31). An increased affinity for MHC could unique microenvironment that allows proper selection (negative drive the more rapid basal homeostatic proliferation of post- and/or positive) of the postrevision TCR repertoire (36). revision T cells. It is unclear why CD4+Vb5+ cells from both nonTg and Vb5Tg It remains to be elucidated whether this potential self-reactivity mice respond poorly to Listeria challenge. It is possible that the is beneficial for arming postrevision T cells with a promiscuous/ Mtv-driven anergy of Vb5+ cells impairs their response. Alterna- cross-reactive TCR repertoire (31) or harmful by giving post- tively, Vb5 TCR chains may lack the ability to recognize the revision cells self-reactive TCR specificities capable of self-tissue LLO190–201 epitope. We cannot distinguish between these possi- destruction. TdT2/2 mice on autoimmune-prone genetic back- bilities experimentally, because H-2b Mtv2 mice derived from grounds are protected from autoimmunity, suggesting that the a wild mouse genetic background (6) display unusual suscepti- decreased N nucleotides of neonatal TCRs protects neonates from bility to Listeria infection (data not shown). generating autoaggressive TCR specificities (32, 33) while en- The proliferation of donor CD4+ Vb5+ and postrevision Vb52 suring functional recognition of foreign Ag with the limited T cells in Rag2/2 recipients was clearly dependent on the pres- available TCR repertoire (31, 34). Similarly, the low number of N ence of bacterial microflora, because antibiotic treatment reduced nucleotides in revised TCRs (18) may allow receptors generated the rate of proliferation and cell recovery for both cell populations through TCR revision to protect the host, absent the functional (Fig. 5). The blunted response to microbial Ags and decreased avidity for self-Ags that can mediate autoimmunity. The absence accumulation of donor Vb5+ cells compared with postrevision of overt autoimmunity in Vb5 Tg mice suggests that postrevision cells likely stems from the anergic state of Vb5+ cells, driven by T cells do not actively mediate autoimmune responses, although chronic encounter with the Mtv tolerogen previous to cell transfer. self-reactive cells may be controlled in these mice by the relatively This well-characterized anergy (4), combined with the weak in- high proportion of Foxp3+ regulatory T cells (21). teraction between Vb5+ T cells and the Mtv ligand (25, 26), also 6534 POSTREVISION T CELLS ARE FUNCTIONAL likely explains the lack of a proliferative response to Mtv by Vb5+ 10. Serra, P., A. Amrani, B. Han, J. Yamanouchi, S. J. Thiessen, and P. Santamaria. + cells transferred to Rag2/2 mice treated with antibiotics. Although 2002. RAG-dependent peripheral T cell receptor diversification in CD8 T lymphocytes. Proc. Natl. Acad. Sci. USA 99: 15566–15571. some donor T cell proliferation occurred even in antibiotic-treated 11. Huang, C. Y., R. Golub, G. E. Wu, and O. Kanagawa. 2002. Superantigen- 2 2 2 Rag / mice, particularly among the Vb5 population (Fig. 5C), induced TCR a locus secondary rearrangement: role in tolerance induction. this likely resulted from the presence of microbes that are in- J. Immunol. 168: 3259–3265. 12. Vaitaitis, G. M., M. Poulin, R. J. Sanderson, K. Haskins, and D. H. Wagner, Jr. sensitive to the antibiotic regimen used. Interestingly, despite the 2003. Cutting edge: CD40-induced expression of recombination activating gene continued (although decreased) proliferation of postrevision cells (RAG) 1 and RAG2: a mechanism for the generation of autoaggressive T cells in in antibiotic-treated Rag2/2 mice (Fig. 5C), their frequency is not the periphery. J. Immunol. 170: 3455–3459. + 13. Lantelme, E., B. Palermo, L. Granziero, S. Mantovani, R. Campanelli, V. Monafo, increased compared with that of Vb5 cells (Fig. 5B), which are A. Lanzavecchia, and C. Giachino. 2000. Cutting edge: recombinase-activating proliferating less extensively. This suggests that the postrevision gene expression and V(D)J recombination in CD4+CD3low mature T lympho- cells in the antibiotic-treated mice may possibly leave secondary cytes. J. Immunol. 164: 3455–3459. 14. Li, T. T., S. Han, M. Cubbage, and B. Zheng. 2002. Continued expression of lymphoid organs and migrate to peripheral tissues, in agreement recombination-activating genes and TCR gene recombination in human pe- with the higher frequency of postrevision cells found in the ripheral T cells. Eur. J. Immunol. 32: 2792–2799. peritoneal lavage compared with the spleens of aged Vb5 Tg mice 15. Lantelme, E., S. Mantovani, B. Palermo, R. Campanelli, L. Granziero, V. Monafo, and C. Giachino. 2000. Increased frequency of RAG-expressing, (data not shown). CD4(+)CD3(low) peripheral T lymphocytes in patients with defective responses Postrevision T cells display TCR-dependent activity for their to DNA damage. Eur. J. Immunol. 30: 1520–1525. maximal proliferation in acutely lymphopenic hosts, proliferation 16. Lantelme, E., V. Turinetto, S. Mantovani, A. Marchi, S. Regazzoni, P. Porcedda, 2/2 M. De Marchi, and C. Giachino. 2003. Analysis of secondary V(D)J rear- to commensal bacterial Ags following transfer to Rag mice, rangements in mature, peripheral T cells of ataxia-telangiectasia heterozygotes. and IFN-g production in response to Listeria infection. It is the- Lab. Invest. 83: 1467–1475. Downloaded from oretically possible that a low level of Vb5 TCR expression on 17. Lantelme, E., L. Orlando, P. Porcedda, V. Turinetto, M. De Marchi, A. Amoroso, S. Mantovani, and C. Giachino. 2008. An in vitro model of T cell receptor re- postrevision T cells is responsible for the TCR signaling in these vision in mature human CD8+ T cells. Mol. Immunol. 45: 328–337. cells. However, this seems unlikely, given that postrevision cells 18. McMahan, C. J., and P. J. Fink. 2000. Receptor revision in peripheral T cells do not proliferate when stimulated with Mtv-overexpressing B creates a diverse V b repertoire. J. Immunol. 165: 6902–6907. 19. Hale, J. S., K. T. Ames, T. E. Boursalian, and P. J. Fink. 2010. Cutting Edge: Rag lymphoma cells. We conclude, therefore, that the TCR-dependent deletion in peripheral T cells blocks TCR revision. J. Immunol. 184: 5964–5968. + 2 + activation/function of postrevision CD4 Vb5 TCRb cells is medi- 20. Hale, J. S., and P. J. Fink. 2010. T-cell receptor revision: friend or foe? Immu- http://www.jimmunol.org/ nology 129: 467–473. ated through the newly generated receptor formed by TCR revi- 21. Zehn, D., M. J. Bevan, and P. J. Fink. 2007. Cutting edge: TCR revision affects sion in the lymphoid periphery. This lack of reactivity to the original predominantly Foxp3 cells and skews them toward the Th17 lineage. J. Immunol. Mtv tolerogen further establishes TCR revision as a means of toler- 179: 5653–5657. 22. Rahemtulla, A., W. P. Fung-Leung, M. W. Schilham, T. M. Ku¨ndig, S. R. Sambhara, ance induction for peripheral T cells that recognize self-Ag. We con- A. Narendran, A. Arabian, A. Wakeham, C. J. Paige, R. M. Zinkernagel, et al. 1991. clude from these cumulative findings that TCR revision acts as a Normal development and function of CD8+ cells but markedly decreased helper cell tolerance mechanism to rescue self-reactive CD4+ T cells, generating activity in mice lacking CD4. Nature 353: 180–184. 23. Grusby, M. J., R. S. Johnson, V. E. Papaioannou, and L. H. Glimcher. 1991. a postrevision T cell population that responds to homeostatic factors + Depletion of CD4 T cells in major histocompatibility complex class II-deficient and possesses immune potential to protect the host against foreign mice. Science 253: 1417–1420. antigenic challenge. 24. Surh, C. D., and J. Sprent. 2008. Homeostasis of naive and memory T cells. by guest on September 23, 2021 Immunity 29: 848–862. 25. Foo-Phillips, M., C. A. Kozak, M. A. C. Principato, and R. Abe. 1992. Char- Acknowledgments acterization of the Mlsf system. II. Identification of mouse mammary tumor virus f We thank Dr. Mark J. Smyth, who generously provided the PN53HI proviruses involved in the clonal deletion of self-Mls -reactive T cells. J. Immunol. 149: 3440–3447. tumor cells, and Drs. Murali-Krishna Kaja and Daniel Stetson for providing 26. Waanders, G. A., R. K. Lees, W. Held, and H. R. MacDonald. 1995. Quantitation mice. of endogenous mouse mammary tumor virus superantigen expression by lym- phocyte subsets. Eur. J. Immunol. 25: 2632–2637. 27. Kieper, W. C., A. Troy, J. T. Burghardt, C. Ramsey, J. Y. Lee, H. Q. Jiang, Disclosures W. Dummer, H. Shen, J. J. Cebra, and C. D. Surh. 2005. Recent immune status The authors have no financial conflicts of interest. determines the source of antigens that drive homeostatic T cell expansion. J. Immunol. 174: 3158–3163. 28. Purton, J. F., J. T. Tan, M. P. Rubinstein, D. M. Kim, J. Sprent, and C. D. Surh. References 2007. Antiviral CD4+ memory T cells are IL-15 dependent. J. Exp. Med. 204: 951–961. 1. Starr, T. K., S. C. Jameson, and K. A. Hogquist. 2003. Positive and negative 29. Feeney, A. J. 1991. Junctional sequences of fetal T cell receptor b chains have selection of T cells. Annu. Rev. Immunol. 21: 139–176. few N regions. J. Exp. Med. 174: 115–124. 2. Seddon, B., and R. Zamoyska. 2003. Regulation of peripheral T-cell homeostasis 30. Bogue, M., S. Cande´ias, C. Benoist, and D. Mathis. 1991. A special repertoire of by receptor signalling. Curr. Opin. Immunol. 15: 321–324. a:b T cells in neonatal mice. EMBO J. 10: 3647–3654. 3. Fink, P. J., K. Swan, G. Turk, M. W. Moore, and F. R. Carbone. 1992. Both 31. Gavin, M. A., and M. J. Bevan. 1995. Increased peptide promiscuity provides intrathymic and peripheral selection modulate the differential expression of V b a rationale for the lack of N regions in the neonatal T cell repertoire. Immunity 3: 5 among CD4+ and CD8+ T cells. J. Exp. Med. 176: 1733–1738. 793–800. 4. Fink, P. J., C. A. Fang, and G. L. Turk. 1994. The induction of peripheral tol- 32. Conde, C., S. Weller, S. Gilfillan, L. Marcellin, T. Martin, and J. L. Pasquali. erance by the chronic activation and deletion of CD4+V b 5+ cells. J. Immunol. 1998. Terminal deoxynucleotidyl transferase deficiency reduces the incidence of 152: 4270–4281. autoimmune nephritis in (New Zealand Black x New Zealand White)F1 mice. 5. McMahan, C. J., and P. J. Fink. 1998. RAG reexpression and DNA re- combination at T cell receptor loci in peripheral CD4+ T cells. Immunity 9: 637– J. Immunol. 161: 7023–7030. 647. 33. Robey, I. F., M. Peterson, M. S. Horwitz, D. H. Kono, T. Stratmann, 6. Blish, C. A., B. J. Gallay, G. L. Turk, K. M. Kline, W. Wheat, and P. J. Fink. A. N. Theofilopoulos, N. Sarvetnick, L. Teyton, and A. J. Feeney. 2004. Terminal 1999. Chronic modulation of the TCR repertoire in the lymphoid periphery. deoxynucleotidyltransferase deficiency decreases autoimmune disease in J. Immunol. 162: 3131–3140. diabetes-prone nonobese diabetic mice and lupus-prone MRL-Fas(lpr) mice. 7. Cooper, C. J., M. T. Orr, C. J. McMahan, and P. J. Fink. 2003. T cell receptor J. Immunol. 172: 4624–4629. revision does not solely target recent thymic emigrants. J. Immunol. 171: 226– 34. Gilfillan, S., M. Bachmann, S. Trembleau, L. Adorini, U. Kalinke, R. Zinkernagel, 233. C. Benoist, and D. Mathis. 1995. Efficient immune responses in mice lacking 8. Takase, M., E. M. Kanagawa, and O. Kanagawa. 2007. Age-dependent TCR N-region diversity. Eur. J. Immunol. 25: 3115–3122. revision mediated by interaction between alphabeta TCR and self-antigens. 35. Gilfillan, S., C. Waltzinger, C. Benoist, and D. Mathis. 1994. More efficient J. Immunol. 179: 2163–2169. positive selection of thymocytes in mice lacking terminal deoxynucleotidyl 9. Bynoe, M. S., C. Viret, R. A. Flavell, and C. A. Janeway, Jr. 2005. T cells from transferase. Int. Immunol. 6: 1681–1686. epicutaneously immunized mice are prone to T cell receptor revision. Proc. Natl. 36. Cooper, C. J., G. L. Turk, M. Sun, A. G. Farr, and P. J. Fink. 2004. Cutting edge: Acad. Sci. USA 102: 2898–2903. TCR revision occurs in germinal centers. J. Immunol. 173: 6532–6536.